{"date_updated":"2023-09-19T14:26:52Z","day":"01","status":"public","language":[{"iso":"eng"}],"_id":"5978","year":"2018","main_file_link":[{"open_access":"1","url":"https://arxiv.org/abs/2004.06370"}],"quality_controlled":"1","publication_status":"published","department":[{"_id":"VlKo"}],"conference":{"start_date":"2018-02-02","name":"AAAI: Conference on Artificial Intelligence","location":"New Orleans, LU, United States","end_date":"2018-02-07"},"type":"conference","publication":"Proceedings of the 32st AAAI Conference on Artificial Intelligence","article_processing_charge":"No","title":"Exact MAP-inference by confining combinatorial search with LP relaxation","page":"6581-6588","oa_version":"Preprint","date_published":"2018-02-01T00:00:00Z","date_created":"2019-02-13T13:32:48Z","isi":1,"scopus_import":"1","abstract":[{"lang":"eng","text":"We consider the MAP-inference problem for graphical models,which is a valued constraint satisfaction problem defined onreal numbers with a natural summation operation. We proposea family of relaxations (different from the famous Sherali-Adams hierarchy), which naturally define lower bounds for itsoptimum. This family always contains a tight relaxation andwe give an algorithm able to find it and therefore, solve theinitial non-relaxed NP-hard problem.The relaxations we consider decompose the original probleminto two non-overlapping parts: an easy LP-tight part and adifficult one. For the latter part a combinatorial solver must beused. As we show in our experiments, in a number of applica-tions the second, difficult part constitutes only a small fractionof the whole problem. This property allows to significantlyreduce the computational time of the combinatorial solver andtherefore solve problems which were out of reach before."}],"publisher":"AAAI Press","user_id":"c635000d-4b10-11ee-a964-aac5a93f6ac1","oa":1,"external_id":{"arxiv":["2004.06370"],"isi":["000485488906082"]},"month":"02","author":[{"last_name":"Haller","first_name":"Stefan","full_name":"Haller, Stefan"},{"first_name":"Paul","last_name":"Swoboda","id":"446560C6-F248-11E8-B48F-1D18A9856A87","full_name":"Swoboda, Paul"},{"full_name":"Savchynskyy, Bogdan","first_name":"Bogdan","last_name":"Savchynskyy"}],"citation":{"short":"S. Haller, P. Swoboda, B. Savchynskyy, in:, Proceedings of the 32st AAAI Conference on Artificial Intelligence, AAAI Press, 2018, pp. 6581–6588.","mla":"Haller, Stefan, et al. “Exact MAP-Inference by Confining Combinatorial Search with LP Relaxation.” Proceedings of the 32st AAAI Conference on Artificial Intelligence, AAAI Press, 2018, pp. 6581–88.","apa":"Haller, S., Swoboda, P., & Savchynskyy, B. (2018). Exact MAP-inference by confining combinatorial search with LP relaxation. In Proceedings of the 32st AAAI Conference on Artificial Intelligence (pp. 6581–6588). New Orleans, LU, United States: AAAI Press.","ista":"Haller S, Swoboda P, Savchynskyy B. 2018. Exact MAP-inference by confining combinatorial search with LP relaxation. Proceedings of the 32st AAAI Conference on Artificial Intelligence. AAAI: Conference on Artificial Intelligence, 6581–6588.","ieee":"S. Haller, P. Swoboda, and B. Savchynskyy, “Exact MAP-inference by confining combinatorial search with LP relaxation,” in Proceedings of the 32st AAAI Conference on Artificial Intelligence, New Orleans, LU, United States, 2018, pp. 6581–6588.","ama":"Haller S, Swoboda P, Savchynskyy B. Exact MAP-inference by confining combinatorial search with LP relaxation. In: Proceedings of the 32st AAAI Conference on Artificial Intelligence. AAAI Press; 2018:6581-6588.","chicago":"Haller, Stefan, Paul Swoboda, and Bogdan Savchynskyy. “Exact MAP-Inference by Confining Combinatorial Search with LP Relaxation.” In Proceedings of the 32st AAAI Conference on Artificial Intelligence, 6581–88. AAAI Press, 2018."}}